Relationship of Abstention Rate between National and Local Elections

Abstract

The two remarkable tendencies have been observed in Japan's recent elections: the continuous rise in the abstention rate at successive elections and the increase of mutohaso (voters who have no particular partisanship). Although many previous studies have devoted attention to voting behavior, the rising abstention rate has been investigated only partially. Additionally, research focused on voting turnout rate have to date tended to slight underlying factors except those affecting the nation as a whole.Nowadays, however, the electorate's distrust of the political system rather than particular parties and statesmen is increasing, and the abstention rate, regardless of whether it is a national or local election, is also rising steadily. This reflects distrust of the political system itself at the local level as well as at the national level. Thus, this rise of abstention rate should be investigated from a more comprehensive framework rather than maintain a bias towards the national level. Therefore, this research note examines the relationship between local elections and the 1996 general election of the House of the Representatives in the southern Kanto region of Japan (Saitama, Chiba, Tokyo and Kanagawa), using the path analysis associated with J. Agnew's concept of context.The abstention rate of both the local and national elections are influenced by the socio-economic characteristics of municipalities located in the region under analysis. However, the high correlation observed between the general election and the two local elections (on municipal and the other prefectural) immediately before it cannot be explained by socio-economic characteristics dlone. This is because the abstention rate at these local elections affected the national election. It is very likely that apathy toward local elections, which had been shared by electors in the context of increasing distrust toward politics in general, caused the result of the higher abstention of the general election through his/her‘path’and‘sense of place’in the terminology of Agnew.Hence, keeping in mind that factors at both the scale of municipalities and prefectures had been an important‘context’for the high abstention of the national election, I applied a causal model which incorporated the impact of the local elections on the national election immediately after them. The results, show a significant causality. It can thus be concluded that Agnew's perspective which enables us to integrate factors working at a few different spatial levels within a single framework is effective for considering phenomena concerned with recent distrust of politics.